LuKreme a écrit :
> I, like everyone I'm sure, have seen a large uptick in spam attempts to
> Message-Ids as if they were email addresses. Of course they are all
> rejected as unknown users, but is it worth putting in a rule to catch
> these specifically?
>
put
reject_unlisted_recipient
reject_unlisted_sender
just after reject_unauth_destination, so that
- you don't run expensive checks (DNSBL, ...) for invalid addresses
- you don't skew your spam blocking stats
a long ago, I used to trap these (redirect to a spamtrap address, do a
PREPEND so that I can add a large score for SA to autolearn them as spam
[well, the real goal is to avoid the opposite], check if the URIs are
listed on uribl and if not, submit them), but I no more do that. I get
enough junk via few other accounts.
I noticed (and others reported similar observations) that if you do this
or if you had a catchall, then you continue getting a lot of junk even
if you disable the traps and the catchall. so think twice before you do
anything here.
